© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 71460
    Prev
    Next
    INDEX
    Explanations

    quoted term definitions

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron is highly responsive to standalone titles or key terms—especially proper names or labels—often presented as section headings or quoted phrases.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    athered
    -1.65
    sistent
    -1.48
    familie
    -1.46
    Szer
    -1.44
    Läs
    -1.42
     eksklu
    -1.41
     bluz
    -1.40
     וכן
    -1.38
    裥
    -1.38
    recently
    -1.36
    POSITIVE LOGITS
     is
    2.88
     was
    2.34
     will
    1.88
     always
    1.66
     usually
    1.66
     being
    1.59
     also
    1.53
     may
    1.50
     now
    1.48
     would
    1.47
    Activations Density 0.035%

    No Known Activations